Present diagnostic methods for active Tuberculosis are inadequate. Acid fast staining and microscopy of sputum is insensitive, particularly in advanced AIDS when atypical smear- negative presentations increase. Mycobacterial culture is the gold standard but is technically demanding and time consuming (i.e., weeks), limiting clinical utility. Novel serologic tests are in development but thus far have been fraught with poor sensitivity and specificity. Tb PCR from sputum is an increasing technology in clinical labs, but has also experienced poor sensitivity in smear-negative specimens. In this proposal we aim to improve Tb PCR sensitivity from sputum with a novel, targeted, DNA extraction methodology. Our hypothesis is that sensitivity of Tb PCR on sputum will increase upon optimal lysis of the lipid-rich tubercle cell wall followed by capture of target Tb sequence with oligonucleotides (thereby eluting the Tb specific DNA away from PCR inhibitory substances in sputum). We have successfully utilized and published on these DNA extraction methods with Giardia qPCR assays in stool, and now propose to adapt them to respiratory specimens. Preliminary data from our group in Tanzania demonstrates the inadequacy of existing diagnostic methods for Tb (smear and culture), shows the utility of Tb PCR on bronchoalveolar lavage fluid, and demonstrates increased PCR sensitivity with DNA capture. In this proposal we will extend this work, analyzing two major methods of Tb DNA extraction. Specific Aim 1 will lyse and capture Tb target DNA utilizing magnetic beads to which Tb-specific oligonucleotides are linked. This method will be adaptable to high-throughput bead- based DNA extraction platforms in U.S. clinical laboratories. Specific Aim 2 will optimize the Tb PCR reaction and assay the final phase I assay in 50 sputum samples in Tanzania. Phase II will plan to take the method to clinical laboratories by adapting the bead assay to high- throughput DNA extraction platforms (e.g., MagNA Pure, Roche) and comparing it against the best-out-there E-MTD test (Gen-Probe). [unreadable] [unreadable] Improved diagnostic methods for smear negative Tuberculosis are unassailably needed. This phase I proposal will test two new targeted Tb DNA extraction methods, a bead-based or acrylamide gel-based capture of lysed Tb DNA, which we hypothesize will increase PCR sensitivity in smear-negative sputa. The most sensitive procedure will be taken in Phase II and adapted to high-throughput DNA extraction platforms and compared against the standard E- MTD assay. [unreadable] [unreadable] [unreadable]